Applied artificial intelligence framework for smart evacuation in industrial disasters

Human evacuation is a critical process during disasters, whether arising from natural events, intentional acts of aggression, or other calamities. The incorporation of diverse computational approaches such as the Internet of Things (IoT) technology and Edge-empowered Cloud platforms has the capability to improve the effectiveness of route recommendation procedures significantly. Conspicuously, this research (i) proposes a sophisticated evacuation framework that integrates the IoT-Edge-Cloud (IEC) computing platform for human evacuation during a disaster; (ii) employs an Artificial Intelligence-based Support Vector Machine (SVM) to detect emergencies in real-time; (iii) facilitates the cloud-based evacuation by computing a safe and swift route using the proposed Markov Decision process. A simulated environment comprising 120,002 data segments is utilized to evaluate the proposed framework. Compared to existing state-of-the-art techniques, improvements in terms of Overall Temporal Delay (37.80 seconds), Energy Efficiency (0.13% per minute), Event Determination Analysis (Accuracy (94.32%)), Route Recommendation Performance (Precision (96.26%), Sensitivity (90.86%), Coverage (96.66%), and Specificity (93.00%)), and Reliability (94.46%) are registered.